What the plans of Greek palaces might have exhibited, did any remains exist, is merely matter for inference and conjecture, and it is not proposed in this volume to pass far beyond ascertained and observed facts. There can be, however, little doubt that the palaces of the West Asiatic style must have at least contributed suggestions as to internal disposition of the later and more magnificent Greek mansions. The ordinary dwelling-houses of citizens, as described by ancient writers, resembled those now visible in the disinterred cities of Pompeii and Herculaneum, which will be referred to under Roman Architecture.[16] The chief characteristic of the plan of these is that they retain the disposition which in the temples was discarded; that is to say, all the doors and windows looked into an inner court, and the house was as far as possible secluded within an encircling wall. The contrast between the openness of the public life led by the men in Greek cities, and the seclusion of the women and the families when at home, is remarkably illustrated by this difference between the public and private buildings. The plan of the triple building called the Erechtheium (Fig. 72) deserves special mention, as an example of an exceptional arrangement which appears to set the ordinary laws of symmetry at defiance, and which is calculated to produce a result into which the picturesque enters at least as much as the beautiful. Though the central temple is symmetrical, the two attached porticoes are not so, and do not, in position, dimensions, or treatment, balance one another. The result is a charming group, and we cannot doubt that other examples of freedom of planning would have been found, had more remains of the architecture of the great cities of Greece come down to our own day. In public buildings other than temples--such as the theatre, the agora, and the basilica--the Greek architects seem to have had great scope for their genius; the planning of the theatres shows skilful and thoroughly complete provisions to meet the requirements of the case. A circular disposition was here introduced--not, it is true, for the first time, since it is rendered probable by the representations on sculptured slabs that some circular buildings existed in Assyria, and circular buildings remain in the archaic works at Mycenæ; but it was now elaborated with remarkable completeness, beauty, and mastery over all the difficulties involved. Could we see the great theatre of Athens as it was when perfect, we should probably find that as an interior it was almost unrivalled, alike for convenience and for beauty; and for these excellences it was mainly indebted to the elegance of its planning. The actual floor of many of the Greek temples appears to have been of marble of different colours. _The Walls._